Page 1 of 2 12 LastLast
Results 1 to 35 of 58
Like Tree7Likes

Thread: Bass Power Amp with GU50

  1. #1
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107

    Bass Power Amp with GU50

    As a member in this forum now for some time - but more reading than contributing - I want to share with you the results of my prototype of a Bass Power Amp created over the last month.

    The target for the design was more than 250W, weight around 20Kg with housing, small size. Here in Germany I have found a company who is delivering customized toroidal transformers for power and output which is a prerequisit to reach the design goal.

    Here are the results up to now, schematic and a picture is in the attachments:
    For the power stage I have selected the GU50, they are reliable and cheap.

    Power at 20Hz: 295W
    Power at 5 KHz: 295W
    Power at 20KHz: 200W
    Weight without housing 15Kg, Size without housing: 53cm x 19cm x 19cm

    Plate voltage at idle is around 950V, at full load around 850V. The operating mode up to now is AB, the next step is to evaluate AB2

    Hans- Georg
    Attached Thumbnails Attached Thumbnails 20120229-bass-power-amp-low-res.jpg  
    Attached Files Attached Files

  2. #2
    Supporting Member Jazz P Bass's Avatar
    Join Date
    Dec 2008
    Location
    Wernersville, PA
    Posts
    11,860
    Wow!
    Does it produce the target 250 watts?

  3. #3
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    As stated: 295Watts starting at 20Hz

  4. #4
    Noodle of Reality Steve Conner's Avatar
    Join Date
    May 2006
    Location
    Glasgow, Scotland
    Posts
    7,251
    Very nice!

    Have you tried playing bass through it? Any thoughts on the sound quality?
    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

  5. #5
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    Yes, I have tested it with my 5 string American Jazzbass using two Reflex boxes with 15 inch speaker.
    From my perception very clean, direct ,with fast response and a lot of headroom.
    Using the active tone control I haven't missed any additional tone control.
    I stopped after playing 10 minutes of full power to save my ears for the future.

  6. #6
    g1
    g1 is offline
    don't forget the joker g1's Avatar
    Join Date
    Feb 2010
    Location
    Canada, somewhere north of Fargo
    Posts
    8,944
    That's a lot of PT secondaries! I was looking at a 300W bass amp that uses a dozen 6L6's (new Traynor), and wondering how your tube costs compare. I only found one seller, but they are only $3.50 each! Are they common or hard to find?
    Certified Dotard

  7. #7
    Senior Member
    Join Date
    Jan 2011
    Location
    Australia
    Posts
    285
    Quote Originally Posted by g-one View Post
    That's a lot of PT secondaries! I was looking at a 300W bass amp that uses a dozen 6L6's (new Traynor), and wondering how your tube costs compare. I only found one seller, but they are only $3.50 each! Are they common or hard to find?
    There's usually tons of russian surplus sellers around for these tubes. The sites are kind of hard to find using google though, unless you know what to look for. One site I frequent had something like 900 of them in stock.

  8. #8
    Senior Member
    Join Date
    Jan 2008
    Location
    Doylestown, PA
    Posts
    1,077
    Awesome! What is your output transformer primary impedance?

    What voltage are the screens running at?

    How are you driving the output grids?

    I ask a lot of questions.

    jamie

  9. #9
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    Hi Jamie,

    here are the answers:

    1. The primary out is set to 2K. I have done some tests with the same OT at 4K (switch the output impedance) and another OT at 2K5, 2K75 and 3K (OT with 3 possible input impedances ). The summary is shown in the diagramm below

    2. The voltage at the screen is 325V at idle mode , about 300V at full output power mode.

    3. I assume that "output grid" is the control grid - g1 - of the GU50.
    Please have a look to the schematic at the top of the thread: the control grids are driven via MOSFET IRF 840, direct coupling.

    Hans- Georg
    Attached Thumbnails Attached Thumbnails output-versus-raa.png  

  10. #10
    Senior Member
    Join Date
    Jan 2008
    Location
    Doylestown, PA
    Posts
    1,077
    Quote Originally Posted by es345 View Post
    Hi Jamie...
    3. I assume that "output grid" is the control grid - g1 - of the GU50.
    Please have a look to the schematic at the top of the thread: the control grids are driven via MOSFET IRF 840, direct coupling.

    Hans- Georg
    Yeah, that's what I meant. I should have used the correct terminology. What is the voltage of the supply rails for the IRF840 relative to ground? (+- voltages)

    Am I correct to assume the phase inverter is capacitively coupled to the mosfet gates (6 of them) which drive the gu50 grids directly and allow for individual bias settings?

    This is a really neat build. I've been wanting to build something like this for a while but haven't yet had the time.

    I think it's interesting to see that the max power was at a slightly higher reflected impedance. How are you measuring the power? (Sine volts rms^2)/load impedance?

    I recently bought a 6.6K:4,8,16 transformer from Edcor with the intent of mismatching it was a 3.3k:2,4,8 transformer for a bass amp. The TF weighs about 12 pounds and should be capable of 200 watts down to around 30Hz. Your amp inspires me to get to work and build something!

    jamie

  11. #11
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    ... if you look to the schematic (pdf file at the beginning of the thread):

    - I have taken one MOSFET for each push pull group to save space taken into account that - as a drawback of that choice - I need then to match the tubes.
    - the MOSFETs are capacitively coupled to the phase inverter, the negative rail for them is - 140V, the positive is +140V. The marging is ok because the bias for that tubes is around -40V to -44V.
    - I have done the measurement with resistive load and sine wave as input. With an oscilloscope I have measured the peak voltage just before the clipping starts.
    - from matching point of view you OT should be OK. Please check the maximum allowed current at the primary. In addition it might be that you will go into saturation at low frequency and full power.

    Hans- Georg

  12. #12
    Senior Member
    Join Date
    Jan 2008
    Location
    Doylestown, PA
    Posts
    1,077
    Yeah, I'm planning to use the Kt120's I've already purchased and I'm shooting for 150 watts. You're right- the output transformer probably couldn't take the current on the bottom end. I am going to start with a few large power transformers I have laying around. I'm tempted to used a Fender Twin Reverb transformer but with bridge rectified secondaries, giving a B+ around a thousand volts. I also have a few old industrial and o-scope PT's giving voltages between 600 and 800 volts. The only issue is that none of them have proper bias windings so I may have to use an addition transformer to generate those voltages, most likely using plus/minus rails and mosfets like you did.

    jamie

  13. #13
    Ted
    Ted is offline
    Senior Member
    Join Date
    May 2007
    Location
    London
    Posts
    293
    That's interesting. I have a large box of GU50s and sockets and I'd been thinking about building something similar.

    Where did you get the transformers? And how much did they cost?

    How did you get on with AB2?

  14. #14
    Noodle of Reality Steve Conner's Avatar
    Join Date
    May 2006
    Location
    Glasgow, Scotland
    Posts
    7,251
    I've got all the parts to build this except the GU50s, want to sell some?
    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

  15. #15
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107

    Smile

    Quote Originally Posted by Ted View Post
    That's interesting. I have a large box of GU50s and sockets and I'd been thinking about building something similar.

    Where did you get the transformers? And how much did they cost?

    How did you get on with AB2?
    The transformers I have got from this company in Germany.
    Ringkerntransformator, Transformatoren, Ringkerntransformatoren | Ringkern, Ringkern-Transformator, 100V-bertrager - BADEL Elektronik GmbH
    They deliver customer defined transformers. The price for both NT and OT together is about 250 Euro.

    Now after tweaking this is the result:

    clean power: 350W (Ub at idle 940V, at full power 840V)
    predicted weight including missing "optical" finish: 22Kg
    to be mentioned: forced cooling by 3 slow speed fans from the bottom through the chassis

    schematic and a picture is included.

    How about the sound?

    I have tested the machine with my 2 EVM15L boxes (vented)
    - with my Jazzbass : unbelievable "pressure", clean fast reaction both in active and passive mode
    - with my ES345 and my Blade basically the same. Obviously I was not trying to test the power stage overdrive...

    In summary: I am satisfied with the result.

    Hans- Georg
    Attached Thumbnails Attached Thumbnails 20121009-brv250-preamp-2-small.jpg   brv-250-rel-3d-power-amp.png   brv-250-rel-3d-power-supply.png  

  16. #16
    Senior Member
    Join Date
    Sep 2010
    Location
    Edmonton AB Canada
    Posts
    257
    Hey glad to hear this amp has carried on into some further design. I won't lie this amp helped me implement my own mosfet driver system, I just have one question about ZD3 and ZD4, are those there to limit the signal swing out of the phase inverter? also the positive voltage to the mosfet followers is only 30 volts is that to limit the amount of positive swing available to drive the tubes grids up? In my amp that rail is around 70V, wonder if i should drop it down?

  17. #17
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    You are right, ZD3/4 are limiting the signal swing. R9a/R10a in series of the Z diodes leads to somewhat more smooth clipping.
    I wanted to be sure not to kill G1 of the powertubes in case of heavy overdrive. If you have 70V at the drain of the mosfets, don't worry. The only difference is a little bit more power dissipation at the mosfets.

  18. #18
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    You are right, ZD3/4 are limiting the signal swing. R9a/R10a in series of the Z diodes leads to somewhat more smooth clipping.
    I wanted to be sure not to kill G1 of the powertubes in case of heavy overdrive. If you have 70V at the drain of the mosfets, don't worry. The only difference is a little bit more power dissipation at the mosfets.

  19. #19
    Senior Member
    Join Date
    Sep 2010
    Location
    Edmonton AB Canada
    Posts
    257
    ok good i figured that was what the zener arrangement was for

  20. #20
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    Here is the last part of the design

    - preamp
    - housing

    Hans- Georg
    Attached Thumbnails Attached Thumbnails brv-250-rel-3d-pre-amp.png   20121014-brv-250-finished-housing-2-small.jpg  

  21. #21
    Ted
    Ted is offline
    Senior Member
    Join Date
    May 2007
    Location
    London
    Posts
    293
    Steve - if you want half a dozen GU50s PM me.

    Hans - do you have specifications for your transformers that you'd be prepared to share?

  22. #22
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    This is what I have ordered (the web page of the company is mentioned above in the thread)

    Both are Toroid types

    OT

    - designed for 350W
    - frequency response from 20 Hz to 20kHz at full power 350W (-3 db limit)
    - Raa 2,2 kOhm; designed for Ub=950 at idle Mode
    - secondary 4 and 8 Ohm

    PT

    Secondary:

    W1 320V 0,1A Preamp+PI
    W2 220V 0,7A Ub Rail 3
    W3 220V 0,7A Ub Rail 2
    W4 240V 0,8A Ub Rail 1 + G2 supply
    W5 100V 0,1A Bias and PI CCS suppy
    W6/W7 6,3V 4A Heater GU50 and MosFET Stage supply
    W8 6,3V 4A Heater Preamp +PI


    If you want to add margin to the GU50 Heater increase W6/W7 to 5A each.
    If you want separate supply for the MosFet stage add W9: 24V 0,2A

    Hans- Georg
    Ted and weiterDrüben like this.

  23. #23
    Senior Member woodyc's Avatar
    Join Date
    Sep 2006
    Posts
    283
    Hi Hans-Georg. Looks like a fun project. I'm having trouble understanding your whole driver/signal limiting strategy. Did AB2 not work out?
    Last edited by woodyc; 12-11-2012 at 08:57 AM.

  24. #24
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    AB2 is working out very well. But you need to take care not to overload G1 of the power tubes as the Mosfets are basically not limiting the current. So you have to control the maximum voltage. What is helpful and what I have implemented is a "smooth clipping" strategy . Some more background you can find here. Grid stoppers with AB2.

  25. #25
    Noodle of Reality Steve Conner's Avatar
    Join Date
    May 2006
    Location
    Glasgow, Scotland
    Posts
    7,251
    I read this thread on DIYAudio: KT88 AB2 Experiences - diyAudio

    Ouch! It confirmed my fears of the worst that could go wrong when driving new production tubes into AB2. GU50 are of course not new production tubes, so they're unlikely to have gold-plated grids that could be damaged.

    My interest in this project is that I have the OPT and driver transformer from a Fender 300PS, and the PT from a FT-DX560 sitting around looking for a use.
    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

  26. #26
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    Reading that thread: really "impressive" to see what can happen to a G1.
    One observation: in the discussed configuration the KT88 are used in pseudotriode mode. In beam pentode mode at the moment of very low anode voltage the high G2 voltage is still "motivating" the electrons to flow rather to the anode than to G1 . This it might protect G1.

    Using my bassamp now since some month the powerstage is quite stable, no drift or loss of power. Obviously I dont play at the power limit but what is amazing is the clean headroom available. (some motivation to start a project )
    Last edited by es345; 12-11-2012 at 08:04 PM. Reason: typo
    Ted and Austin like this.

  27. #27
    Senior Member woodyc's Avatar
    Join Date
    Sep 2006
    Posts
    283
    Quote Originally Posted by es345 View Post
    AB2 is working out very well...
    Well I was looking at some GU50 characteristics for VG2=250 and it looks like you could get about 310W with a 900V B+, a primary of around 3500 Ohms and no grid current. Since you have 50 more volts on the screens, it seems like something very close to 350W should be possible with no grid current. If you are getting anything from grid current it must only be a small fraction of a dB?

    gu50.gif

  28. #28
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    taking your diagram you have a saturation voltage at the anode of about 150V.
    If you have a OT with a coil resistance center to anode of about 70 Ohm (this is my case) you will get about 270 W output, not more. (see the diagram below)

    Now my case:
    If you look to the beginning of the thread you will find that I have reached 300W without AB2.With AB2 it is now somewhere between 350 and 360, so I just count it for 350W.
    This is ~0,7 db and that is what I have expected.

    Why not more?
    AB2 helps you in bringing down the saturation voltage, something like 50V. If you have high anode voltage the possible gain is limited.

    An example:
    - UB=900, Uanode min =150V >> U peak at OT, center to anode=750V
    - now you go to AB2 reducing Uanode min to 100V >> Upeak at OT center to anode= 800V
    this is about 0,56db more.

    So is it worth to do?

    For me the decision was yes, because I definitely haven't any crossover distortion (Prio 1) and i have some watt more output (Prio 2). And not to forget: It was a great fun to develop and learn something about AB2.
    Attached Thumbnails Attached Thumbnails 6xgu50-3k5-900v.jpg  

  29. #29
    Senior Member jazbo8's Avatar
    Join Date
    Jun 2011
    Posts
    599
    Very impressive build! Would it be possible for you to share your Excel worksheet? Looks like a handy tool to have...

    TIA,
    Jaz

  30. #30
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    you can find it here Smartest way to 200W? - Page 2

  31. #31
    Senior Member jazbo8's Avatar
    Join Date
    Jun 2011
    Posts
    599
    Quote Originally Posted by es345 View Post
    you can find it here Smartest way to 200W? - Page 2
    Thanks! Now I gotta read that thread too...

  32. #32
    Noodle of Reality Steve Conner's Avatar
    Join Date
    May 2006
    Location
    Glasgow, Scotland
    Posts
    7,251
    Another advantage of AB2 is that the power stage can be overdriven without charging up the coupling capacitors and getting crossover distortion. That's part of the sound of many classic guitar amps, but it's maybe not so hot for bass.

    The Ampeg SVT used a diode limiter like es345's to avoid this. I proposed using a Vactrol compressor with the LED driven by the grid current (or a filament lamp driven by screen current)
    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

  33. #33
    Senior Member woodyc's Avatar
    Join Date
    Sep 2006
    Posts
    283
    Quote Originally Posted by es345 View Post
    ...If you have a OT with a coil resistance center to anode of about 70 Ohm (this is my case) you will get about 270 W output, not more...
    I guess that begs the question: why 70 Ohms? It seems like a lot of power to leave in your transformer.

    For a practical example, I had an ampeg V4 on my bench yesterday. The center tap to anode measures lower than yours (54 Ohms dc) -- an thats only a 120W amp!

    By the way, I don't mean to sound like I'm complaining. I'm just trying to understand your design choices. Your project is very impressive and inspiring.

  34. #34
    Senior Member
    Join Date
    Mar 2011
    Location
    Ulm, Germany
    Posts
    107
    No problem at all

    Well thats what I have measured and it has told me something:

    1. Don't forget that thing when you estimate the output power.
    2. The next time I will put a low Raa resistance to the specifications.

    There might be one other reason for that high value: I have defined a maximum size of the OT...

  35. #35
    Noodle of Reality Steve Conner's Avatar
    Join Date
    May 2006
    Location
    Glasgow, Scotland
    Posts
    7,251
    Yes, the physics of transformers won't allow you to have small size, low DC resistance and extended bass response all together.

    As far as I know, the 300PS OT is only specified down to 50Hz at full power.
    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

Page 1 of 2 12 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. GU50 in a bass or guitar amp?
    By imaradiostar in forum Theory & Design
    Replies: 24
    Last Post: 10-30-2011, 12:10 PM
  2. Advice on Hartke 3500 bass amp repair--power amp module
    By drm1 in forum Maintenance, Troubleshooting & Repair
    Replies: 12
    Last Post: 04-18-2011, 05:20 PM
  3. Schematic critique: GU50 SE amp
    By tubby in forum Theory & Design
    Replies: 4
    Last Post: 02-23-2010, 02:53 PM
  4. Bass amp to Power amp conversion? (70's Yamaha B100II)
    By jaywalker512 in forum Mods & Tweaks
    Replies: 13
    Last Post: 12-08-2009, 12:37 AM
  5. High power bass amp
    By maciek in forum Theory & Design
    Replies: 6
    Last Post: 08-17-2009, 01:29 PM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •